Heat Index: LTC de Meent

This repository contains a small single-purpose web application intended to show heat-index data for LTC de Meent in Bussum: the current value, history based on available data, and forecast values.

The intended primary data source is KNMI:

• Historical data.
• Current data.
• Forecasts.

Until a KNMI API key is available, the published application uses temporary no-key sources:

• Current observed conditions for comparison: Buienradar public station feed, https://data.buienradar.nl/2.0/feed/json.
• Forecast and temporary grid backfill data: Open-Meteo Forecast API, https://open-meteo.com/en/docs.
• Open-Meteo is queried by latitude/longitude for temperature_2m and relative_humidity_2m, with past_days=7 and forecast_days=8.
• Hourly Open-Meteo values are interpolated into 30-minute point values.
• The JSON contains 721 point values: 15 days of half-hour intervals plus the final midnight endpoint.
• Each 30-minute heatmap cell represents the rounded point-in-time timestamp shown for that cell. For example, the 14:30 cell shows the interpolated value at 14:30.
• Nearby Buienradar station observations are stored as comparison data only; they do not overwrite the Open-Meteo grid value.

All data is placed as close to: 52.27708° N, 5.14383° E

Tech Stack

Single-page web application, with minimum dependencies:

• No build-time required: resources are as-is.
• Bootstrap CSS is loaded from CDN.
• Any other dependencies need to be considered.

The application is published to GitHub Pages via CI/CD, attached to the public deployment environment.

Data Generation

Generate live temporary data from Open-Meteo with Node.js LTS:

node scripts/refresh.mjs --output public/data/heat-index.json --history public/data/heat-index-history.json

Generate local sample data without network access:

node scripts/refresh.mjs --sample --output public/data/heat-index.json --history public/data/heat-index-history.json

The refresh step writes:

• public/data/heat-index.json: the current 15-day display window.
• public/data/heat-index-history.json: accumulated actual point values, forecast runs, and forecast-change records detected between refreshes.

The generated files in public/data/*.json are committed so published history and current values can be reviewed in Git.

Heat Index Formula

The application calculates heat index from air temperature and relative humidity using the NOAA/NWS Rothfusz regression. This is the standard heat-index equation published by the U.S. National Weather Service and is intended to match the temperature/humidity heat-index model used by services such as KNLTB and weerplaza.nl.

Authoritative reference: https://www.wpc.ncep.noaa.gov/html/heatindex_equation.shtml

Inputs:

• T: air temperature in degrees Fahrenheit.
• RH: relative humidity as a percentage.

Main equation:

$$ \begin{aligned} HI ={}& -42.379 + 2.04901523T + 10.14333127RH \\ &- 0.22475541T RH - 0.00683783T^2 - 0.05481717RH^2 \\ &+ 0.00122874T^2 RH + 0.00085282T RH^2 - 0.00000199T^2 RH^2 \end{aligned} $$

For cooler conditions, the NWS simple heat-index approximation is used first:

$$ HI = 0.5 \times \left(T + 61.0 + \left((T - 68.0) \times 1.2\right) + (RH \times 0.094)\right) $$

If the average of this approximation and the actual air temperature is below 80 °F, that value is used. Otherwise the Rothfusz regression is used, including the NWS low-humidity and high-humidity adjustments:

Low-humidity adjustment:

$$ \frac{13 - RH}{4} \times \sqrt{\frac{17 - |T - 95|}{17}} $$

Applied when:

$$ RH < 13 \quad\text{and}\quad 80 \le T \le 112 $$

High-humidity adjustment:

$$ \frac{RH - 85}{10} \times \frac{87 - T}{5} $$

Applied when:

$$ RH > 85 \quad\text{and}\quad 80 \le T \le 87 $$

The calculation is performed in Fahrenheit and converted back to Celsius for display. Heat index is a shade-based apparent temperature. Direct sunlight can make conditions feel hotter than the displayed value.